In September, autonomous vehicles at level 3, if the driver’s attention is no longer required, will be able to circulate on our roads.
In September, autonomous vehicles at level 3, if the driver’s attention is no longer required, will be able to circulate on our roads. Road traffic is not the only area where autonomous driving has strong economic and societal challenges. The industry and the management of its production lines are also undergoing profound changes that lead us to automatic piloting: the autonomous factory.
Support change management…
Vehicle autonomy is leading to rethinking the future of driver training. Soon the driver will only have to regain control of the vehicle in complex or emergency situations. Like an airplane pilot, it will be more about piloting than driving. In the world of industry, the question of education is quite another. Gone are the long careers where the worker is a world-class pilot, able to recognize by ear the correct setting for the machine. Industrial production must adapt to the new generation: go from piloting to a few hours of training for autonomous and productive traffic on production lines.
However, technology does not solve the problem of individual and societal trust. Our sense of trust in third-party driving largely depends on how close the driving behavior is to our own behavior. If the vehicle learns from our behavior, we also learn to trust the built-in algorithm as we go. Limiter, cruise control, autonomous parking, lane keeping and semi-autonomous driving on main roads… All these systems transform our relationship with the vehicle and build our confidence. Without this, it is a safe bet that the transition to autonomous vehicles would only go through a generational renewal of drivers.
The parallel to the autonomous factory is obvious. A team that has been running a production line for years has no reason to trust an algorithm and have its decision making challenged or sidelined. Trust must be built through a step-by-step transition. An Industry 4.0 project that does not involve teams and change management is doomed to failure.
… by proceeding step by step
An independent factory project starts, as for the car, with specific actions, such as situation analysis or solving specific problems. These solutions, whose results come from data – and therefore from the past in the production chain – are explainable, contextualized and based on previous choices made by the teams. Trust can then begin to be established through concrete business results.
The next step is to go from the punctual to the daily. The one-time analysis does not make it possible to react to dangers. It then becomes necessary to integrate the analysis and recommendation into production in real time. The trust established in the previous step is therefore necessary to reach this level of maturity. The use of prescriptive recommendations, that is, based directly on previous and justified production experiences, is a key to acceptance.
The third level of the autonomous factory consists in automating the decision-making of the application of dynamic instructions proposed by a prescriptive system. The factory then becomes autonomous, and the worker assists the system, for example when unknown situations arise. Just as our vehicle will first be autonomous on dry roads by giving us control in the rain before we can handle all situations.
The final stage of transformation occurs by complementing the prescriptive approach with predictive decision making. The control algorithm will enrich its AI with a system capable of suggesting configurations, even in situations that it has never been exposed to. A situation that requires full confidence in the implemented solution. This transformation towards the autonomous factory is therefore a long road to be taken while being fully aware of the steps to be taken.
In addition to addressing economic and ecological issues, the autonomous factory frees up time and allows teams to focus on value-creating tasks. It thus leads to the creation of new professions, which in turn will supply the technology with new knowledge as the process evolves.